library(tidyverse)
## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.0 ──
## ✓ ggplot2 3.3.2 ✓ purrr 0.3.4
## ✓ tibble 3.0.4 ✓ dplyr 1.0.2
## ✓ tidyr 1.1.2 ✓ stringr 1.4.0
## ✓ readr 1.4.0 ✓ forcats 0.5.0
## Warning: package 'ggplot2' was built under R version 3.6.2
## Warning: package 'tibble' was built under R version 3.6.2
## Warning: package 'tidyr' was built under R version 3.6.2
## Warning: package 'readr' was built under R version 3.6.2
## Warning: package 'purrr' was built under R version 3.6.2
## Warning: package 'dplyr' was built under R version 3.6.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
library(readr)
library(coefplot)
Homeowner_data <- read_csv("~/Desktop/Econometrics/Homeowner.data.csv")
##
## ── Column specification ────────────────────────────────────────────────────────
## cols(
## .default = col_double(),
## AGE_REF_ = col_character(),
## AGE2_ = col_character(),
## CUTE_URE = col_character(),
## DESCRIP_ = col_character(),
## EARN_OMP = col_character(),
## EDUC0REF = col_character(),
## EDUCA2_ = col_character(),
## EMPL_YP1 = col_character(),
## EMPL_YP2 = col_character(),
## FAM__IZE = col_character(),
## FAM__YPE = col_character(),
## FGVX_ = col_character(),
## FINC_EFX = col_character(),
## FIRAX_ = col_character(),
## FJSS_EDX = col_character(),
## FPVTX_ = col_character(),
## FREEMLX_ = col_character(),
## FRRX_ = col_character(),
## FS_MTHI_ = col_character(),
## FSS_RRX_ = col_character()
## # ... with 78 more columns
## )
## ℹ Use `spec()` for the full column specifications.
## Warning: 36 parsing failures.
## row col expected actual file
## 3798 ROYESTB 1/0/T/F/TRUE/FALSE 12 '~/Desktop/Econometrics/Homeowner.data.csv'
## 3798 ROYESTBX 1/0/T/F/TRUE/FALSE 60000 '~/Desktop/Econometrics/Homeowner.data.csv'
## 3799 ROYESTB 1/0/T/F/TRUE/FALSE 12 '~/Desktop/Econometrics/Homeowner.data.csv'
## 3799 ROYESTBX 1/0/T/F/TRUE/FALSE 60000 '~/Desktop/Econometrics/Homeowner.data.csv'
## 4233 ROYESTB 1/0/T/F/TRUE/FALSE 8 '~/Desktop/Econometrics/Homeowner.data.csv'
## .... ........ .................. ...... ...........................................
## See problems(...) for more details.
#The code below creates my subset of the data which is between and including the ages of 20 and 87 with a family size of at least 3.
#I also setup RACE, HOMEOWNERSHIP, MARITAL STATUS, & EDUC level as factors (some ordered).
Homeowner_data$REF_RACE <- as.factor(Homeowner_data$REF_RACE)
levels(Homeowner_data$REF_RACE) <- c("White", "Black", "Native American", "Asian", "Pacific Islander", "Multi Race")
Homeowner_data$CUTENURE <- as.factor(Homeowner_data$CUTENURE)
levels(Homeowner_data$CUTENURE) <- c("Mortgage","No Mortgage","Mortgage Status Not Reported", "Renter", "Occupied Without Payment Of Cash", "Student Housing")
Homeowner_data$EDUC_REF <- as.factor(Homeowner_data$EDUC_REF)
levels(Homeowner_data$EDUC_REF) <- c("Never Attended","Grade 8","High School No Degree", "High School", "Some College", "Associates", "Bachelors", "Masters or PHD")
Homeowner_data$MARITAL1 <- as.factor(Homeowner_data$MARITAL1)
levels(Homeowner_data$MARITAL1) <- c("Married","Widowed","Divorced", "Seperated", "Never Married")
use_varb <- (Homeowner_data$AGE_REF >= 20) & (Homeowner_data$AGE_REF <= 87) & (Homeowner_data$FAM_SIZE >= 3)
dat_use <- subset(Homeowner_data,use_varb)
#The regression below attempts to find a relationship between yearly wages and several variables in order to find what variables in the data set would be good predictors for a male homeowner in my subset.
model_temp1 <- lm(FWAGEXM ~ AGE_REF + EDUC_REF + REF_RACE + CUTENURE + SEX_REF, data = dat_use)
summary(model_temp1)
##
## Call:
## lm(formula = FWAGEXM ~ AGE_REF + EDUC_REF + REF_RACE + CUTENURE +
## SEX_REF, data = dat_use)
##
## Residuals:
## Min 1Q Median 3Q Max
## -177426 -41545 -10907 26402 525304
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 97041.90 18145.25 5.348 9.43e-08
## AGE_REF 180.09 97.64 1.844 0.06520
## EDUC_REFGrade 8 -18862.21 18330.06 -1.029 0.30353
## EDUC_REFHigh School No Degree -7674.52 17774.13 -0.432 0.66593
## EDUC_REFHigh School 1098.07 17420.02 0.063 0.94974
## EDUC_REFSome College 6979.04 17439.30 0.400 0.68904
## EDUC_REFAssociates 11213.93 17602.20 0.637 0.52411
## EDUC_REFBachelors 49129.13 17424.48 2.820 0.00483
## EDUC_REFMasters or PHD 75711.93 17528.82 4.319 1.61e-05
## REF_RACEBlack -11959.95 4180.51 -2.861 0.00425
## REF_RACENative American -5325.03 11739.80 -0.454 0.65015
## REF_RACEAsian 6543.54 4249.44 1.540 0.12368
## REF_RACEPacific Islander 4959.27 16384.94 0.303 0.76216
## REF_RACEMulti Race 21443.21 7785.16 2.754 0.00591
## CUTENURENo Mortgage -36966.78 3600.33 -10.268 < 2e-16
## CUTENUREMortgage Status Not Reported -16900.06 10300.98 -1.641 0.10096
## CUTENURERenter -40109.96 2793.49 -14.358 < 2e-16
## CUTENUREOccupied Without Payment Of Cash -39552.95 12313.63 -3.212 0.00133
## SEX_REF -10335.62 2302.86 -4.488 7.40e-06
##
## (Intercept) ***
## AGE_REF .
## EDUC_REFGrade 8
## EDUC_REFHigh School No Degree
## EDUC_REFHigh School
## EDUC_REFSome College
## EDUC_REFAssociates
## EDUC_REFBachelors **
## EDUC_REFMasters or PHD ***
## REF_RACEBlack **
## REF_RACENative American
## REF_RACEAsian
## REF_RACEPacific Islander
## REF_RACEMulti Race **
## CUTENURENo Mortgage ***
## CUTENUREMortgage Status Not Reported
## CUTENURERenter ***
## CUTENUREOccupied Without Payment Of Cash **
## SEX_REF ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 68830 on 3715 degrees of freedom
## Multiple R-squared: 0.255, Adjusted R-squared: 0.2514
## F-statistic: 70.66 on 18 and 3715 DF, p-value: < 2.2e-16
plot(model_temp1)
require(stargazer)
## Loading required package: stargazer
##
## Please cite as:
## Hlavac, Marek (2018). stargazer: Well-Formatted Regression and Summary Statistics Tables.
## R package version 5.2.2. https://CRAN.R-project.org/package=stargazer
stargazer(model_temp1, type = "default")
##
## % Error: 'style' must be either 'latex' (default), 'html' or 'text.'
# The strong predictors for male homeownership seems to be wages, education, race. When family size is added to the subset the wages for both asian male and black male homeowners the slope becomes positive. Do extra family members persuade companies to pay more in wages? Something to think about.
reg_biv <- lm(FWAGEXM ~ AGE_REF + REF_RACE + EDUC_REF, data = dat_use)
age_35_bachelors <- coef(reg_biv)[1] + 35*coef(reg_biv)[2] + coef(reg_biv)[3] + coef(reg_biv)[4]
summary(reg_biv)
##
## Call:
## lm(formula = FWAGEXM ~ AGE_REF + REF_RACE + EDUC_REF, data = dat_use)
##
## Residuals:
## Min 1Q Median 3Q Max
## -177694 -44356 -10379 28879 530036
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 47864.7 18292.0 2.617 0.008914 **
## AGE_REF 306.8 95.7 3.206 0.001356 **
## REF_RACEBlack -18502.4 4297.1 -4.306 1.71e-05 ***
## REF_RACENative American -13787.7 12163.4 -1.134 0.257061
## REF_RACEAsian 2379.3 4385.5 0.543 0.587470
## REF_RACEPacific Islander -13840.7 16954.4 -0.816 0.414355
## REF_RACEMulti Race 16863.2 8064.4 2.091 0.036590 *
## EDUC_REFGrade 8 -20237.8 19009.8 -1.065 0.287126
## EDUC_REFHigh School No Degree -6600.3 18432.9 -0.358 0.720309
## EDUC_REFHigh School 7322.9 18055.3 0.406 0.685073
## EDUC_REFSome College 18883.2 18054.8 1.046 0.295684
## EDUC_REFAssociates 25387.5 18217.2 1.394 0.163520
## EDUC_REFBachelors 65658.5 18021.9 3.643 0.000273 ***
## EDUC_REFMasters or PHD 93021.1 18129.0 5.131 3.03e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 71410 on 3720 degrees of freedom
## Multiple R-squared: 0.197, Adjusted R-squared: 0.1942
## F-statistic: 70.2 on 13 and 3720 DF, p-value: < 2.2e-16
print("Mean Wage of Age 35 person in my subset")
## [1] "Mean Wage of Age 35 person in my subset"
print(age_35_bachelors)
## (Intercept)
## 26314
#Predicting the peak wage age of a black male homeowner with a bachelors degree.
NNobs <- length(dat_use$FWAGEXM)
set.seed(12345)
graph_obs <- (runif(NNobs) < 0.1)
dat_graph <-subset(dat_use,graph_obs)
plot(FWAGEXM ~ jitter(AGE_REF, factor = 2), pch = 16, col = rgb(1, 0.2, 0.6, alpha = 0.2), main = "Wage vs Age for a black female homeowner with a bachelors degree", xlab = "Age", ylab = "Wage", ylim = c(40000,150000), data = dat_graph)
to_be_predicted2 <- data.frame(AGE_REF = 20:87, REF_RACE = "Black", EDUC_REF = "Bachelors", CUTENURE = "Mortgage", CUTENURE = "No Mortgage", CUTENURE = "Mortgage Status Not Reported", SEX_REF = 2)
to_be_predicted2$yhat <- predict(model_temp1, newdata = to_be_predicted2)
lines(yhat ~ AGE_REF, data = to_be_predicted2)
# For a black male homeowner with a bachelors degree, yearly wage starts at around $130,000 at age 20 and decreases at all ages from 20 to 80.
#Predicting the peak wage age of an asian male homeowner with a bachelors degree.
NNobs <- length(dat_use$FWAGEXM)
set.seed(12345)
graph_obs <- (runif(NNobs) < 0.1)
dat_graph <-subset(dat_use,graph_obs)
plot(FWAGEXM ~ jitter(AGE_REF, factor = 2), pch = 16, col = rgb(1, 0.2, 0.6, alpha = 0.2), main = "Wage vs Age for an asian female homeowner with a bachelors degree", xlab = "Age", ylab = "Wage", ylim = c(80000,170000), data = dat_graph)
to_be_predicted2 <- data.frame(AGE_REF = 20:87, REF_RACE = "Asian", EDUC_REF = "Bachelors", CUTENURE = "Mortgage", CUTENURE = "No Mortgage", CUTENURE = "Mortgage Status Not Reported", SEX_REF = 2)
to_be_predicted2$yhat <- predict(model_temp1, newdata = to_be_predicted2)
lines(yhat ~ AGE_REF, data = to_be_predicted2)
# For an asian male homeowner with a bachelors degree, yearly wage starts at around $170,000 at age 20 and decreases at all ages from 20 to 80 but the slope is not as high (negatively) compared to the black male.
#This is interesting because it shows that black men are making less wages at all ages vs asian men with the same college degree.
#When family size is added to the subset the graph for wages for both asian male and black male homeowners has a positive slope. Do extra family members persuade companies to pay more in wages? Maybe family men are more likely to strive for higher wages? Something to think about.
# taking the log of the wage function allows for comparing values using percent changes and reducing the effect of education on wage.
model_temp3 <- lm(log1p(FWAGEXM) ~ AGE_REF + REF_RACE + EDUC_REF + CUTENURE + MARITAL1 + SEX_REF, data = dat_use)
summary(model_temp3)
##
## Call:
## lm(formula = log1p(FWAGEXM) ~ AGE_REF + REF_RACE + EDUC_REF +
## CUTENURE + MARITAL1 + SEX_REF, data = dat_use)
##
## Residuals:
## Min 1Q Median 3Q Max
## -11.9607 -0.1657 0.3955 0.9879 3.8799
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 12.922529 0.613073 21.078 < 2e-16
## AGE_REF -0.024107 0.003446 -6.996 3.10e-12
## REF_RACEBlack -0.108472 0.141792 -0.765 0.444317
## REF_RACENative American -0.913539 0.395759 -2.308 0.021036
## REF_RACEAsian -0.006162 0.143133 -0.043 0.965661
## REF_RACEPacific Islander -1.554725 0.551746 -2.818 0.004861
## REF_RACEMulti Race 0.337784 0.262248 1.288 0.197815
## EDUC_REFGrade 8 -1.349111 0.616942 -2.187 0.028822
## EDUC_REFHigh School No Degree -1.009188 0.597265 -1.690 0.091173
## EDUC_REFHigh School -0.569652 0.586017 -0.972 0.331076
## EDUC_REFSome College -0.497554 0.586568 -0.848 0.396356
## EDUC_REFAssociates -0.108913 0.592013 -0.184 0.854047
## EDUC_REFBachelors 0.124732 0.586284 0.213 0.831534
## EDUC_REFMasters or PHD 0.450917 0.589770 0.765 0.444580
## CUTENURENo Mortgage -1.018139 0.121578 -8.374 < 2e-16
## CUTENUREMortgage Status Not Reported -0.561157 0.346200 -1.621 0.105124
## CUTENURERenter -0.685695 0.095480 -7.182 8.29e-13
## CUTENUREOccupied Without Payment Of Cash -0.116183 0.414628 -0.280 0.779332
## MARITAL1Widowed -0.672777 0.258321 -2.604 0.009240
## MARITAL1Divorced -0.340558 0.139863 -2.435 0.014941
## MARITAL1Seperated -0.550229 0.241333 -2.280 0.022667
## MARITAL1Never Married -0.753885 0.134014 -5.625 1.99e-08
## SEX_REF -0.273525 0.078652 -3.478 0.000512
##
## (Intercept) ***
## AGE_REF ***
## REF_RACEBlack
## REF_RACENative American *
## REF_RACEAsian
## REF_RACEPacific Islander **
## REF_RACEMulti Race
## EDUC_REFGrade 8 *
## EDUC_REFHigh School No Degree .
## EDUC_REFHigh School
## EDUC_REFSome College
## EDUC_REFAssociates
## EDUC_REFBachelors
## EDUC_REFMasters or PHD
## CUTENURENo Mortgage ***
## CUTENUREMortgage Status Not Reported
## CUTENURERenter ***
## CUTENUREOccupied Without Payment Of Cash
## MARITAL1Widowed **
## MARITAL1Divorced *
## MARITAL1Seperated *
## MARITAL1Never Married ***
## SEX_REF ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 2.312 on 3711 degrees of freedom
## Multiple R-squared: 0.1232, Adjusted R-squared: 0.118
## F-statistic: 23.7 on 22 and 3711 DF, p-value: < 2.2e-16
plot(model_temp3)
require(stargazer)
stargazer(model_temp3, type = "default")
##
## % Error: 'style' must be either 'latex' (default), 'html' or 'text.'
REF_RACEB <- factor(c("0", "1", "0", "0", "0", "0"))
as.logical(as.integer(levels(REF_RACEB)[REF_RACEB]))
## [1] FALSE TRUE FALSE FALSE FALSE FALSE
as.logical(as.integer(REF_RACEB) - 1L)
## [1] FALSE TRUE FALSE FALSE FALSE FALSE
as.logical(as.integer(as.character(REF_RACEB)))
## [1] FALSE TRUE FALSE FALSE FALSE FALSE
as.logical(REF_RACEB)
## [1] NA NA NA NA NA NA
levels(REF_RACEB) <- c(FALSE,TRUE)
REF_RACEB <- as.logical(REF_RACEB)
na.omit(REF_RACEB)
## [1] FALSE TRUE FALSE FALSE FALSE FALSE
SEX_REFB <- factor(c("0", "1"))
as.logical(as.integer(levels(SEX_REFB)[SEX_REFB]))
## [1] FALSE TRUE
as.logical(as.integer(SEX_REFB) - 1L)
## [1] FALSE TRUE
as.logical(as.integer(as.character(SEX_REFB)))
## [1] FALSE TRUE
levels(SEX_REFB) <- c(FALSE,TRUE)
SEX_REFB <- as.logical(SEX_REFB)
na.omit(SEX_REFB)
## [1] FALSE TRUE
EDUC_REFB <- factor(c("0", "0", "0", "1", "0", "0", "0", "0"))
as.logical(as.integer(levels(EDUC_REFB)[EDUC_REFB]))
## [1] FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE
as.logical(as.integer(EDUC_REFB) - 1L)
## [1] FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE
as.logical(as.integer(as.character(EDUC_REFB)))
## [1] FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE
levels(EDUC_REFB) <- c(FALSE,TRUE)
EDUC_REFB <- as.logical(EDUC_REFB)
na.omit(EDUC_REFB)
## [1] FALSE FALSE FALSE TRUE FALSE FALSE FALSE FALSE
model_temp2 <- lm(FWAGEXM ~ AGE_REF + I(AGE_REF^2) + I(REF_RACEB * SEX_REF) + SEX_REF + REF_RACE + EDUC_REF + CUTENURE + MARITAL1, data = dat_use)
## Warning in REF_RACEB * SEX_REF: longer object length is not a multiple of
## shorter object length
summary(model_temp2)
##
## Call:
## lm(formula = FWAGEXM ~ AGE_REF + I(AGE_REF^2) + I(REF_RACEB *
## SEX_REF) + SEX_REF + REF_RACE + EDUC_REF + CUTENURE + MARITAL1,
## data = dat_use)
##
## Residuals:
## Min 1Q Median 3Q Max
## -183320 -40797 -9609 27303 513902
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) -33834.200 21708.354 -1.559 0.119181
## AGE_REF 6463.718 589.986 10.956 < 2e-16
## I(AGE_REF^2) -66.044 6.152 -10.735 < 2e-16
## I(REF_RACEB * SEX_REF) -1753.541 1831.562 -0.957 0.338427
## SEX_REF -6785.838 2293.716 -2.958 0.003111
## REF_RACEBlack -9596.594 4110.661 -2.335 0.019619
## REF_RACENative American -3739.446 11471.742 -0.326 0.744465
## REF_RACEAsian 3830.872 4151.552 0.923 0.356194
## REF_RACEPacific Islander 3923.299 15995.658 0.245 0.806259
## REF_RACEMulti Race 26067.047 7611.269 3.425 0.000622
## EDUC_REFGrade 8 -36112.116 17892.346 -2.018 0.043632
## EDUC_REFHigh School No Degree -15203.739 17316.309 -0.878 0.380000
## EDUC_REFHigh School -7907.180 16987.902 -0.465 0.641630
## EDUC_REFSome College -2772.606 17005.168 -0.163 0.870492
## EDUC_REFAssociates 1487.724 17164.122 0.087 0.930933
## EDUC_REFBachelors 36271.987 17000.472 2.134 0.032942
## EDUC_REFMasters or PHD 62204.465 17103.112 3.637 0.000280
## CUTENURENo Mortgage -28347.694 3564.162 -7.954 2.39e-15
## CUTENUREMortgage Status Not Reported -16378.927 10032.770 -1.633 0.102650
## CUTENURERenter -32315.659 2778.991 -11.629 < 2e-16
## CUTENUREOccupied Without Payment Of Cash -26638.492 12037.567 -2.213 0.026962
## MARITAL1Widowed -22959.211 7528.770 -3.050 0.002308
## MARITAL1Divorced -31122.917 4069.526 -7.648 2.59e-14
## MARITAL1Seperated -36613.101 6993.784 -5.235 1.74e-07
## MARITAL1Never Married -12741.408 3969.143 -3.210 0.001338
##
## (Intercept)
## AGE_REF ***
## I(AGE_REF^2) ***
## I(REF_RACEB * SEX_REF)
## SEX_REF **
## REF_RACEBlack *
## REF_RACENative American
## REF_RACEAsian
## REF_RACEPacific Islander
## REF_RACEMulti Race ***
## EDUC_REFGrade 8 *
## EDUC_REFHigh School No Degree
## EDUC_REFHigh School
## EDUC_REFSome College
## EDUC_REFAssociates
## EDUC_REFBachelors *
## EDUC_REFMasters or PHD ***
## CUTENURENo Mortgage ***
## CUTENUREMortgage Status Not Reported
## CUTENURERenter ***
## CUTENUREOccupied Without Payment Of Cash *
## MARITAL1Widowed **
## MARITAL1Divorced ***
## MARITAL1Seperated ***
## MARITAL1Never Married **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 66990 on 3709 degrees of freedom
## Multiple R-squared: 0.2954, Adjusted R-squared: 0.2908
## F-statistic: 64.78 on 24 and 3709 DF, p-value: < 2.2e-16
plot(model_temp2)
require(stargazer)
stargazer(model_temp2, type = "default")
##
## % Error: 'style' must be either 'latex' (default), 'html' or 'text.'
pick_use1 <- (Homeowner_data$AGE_REF >= 20) & (Homeowner_data$AGE_REF <= 87) & (Homeowner_data$FAM_SIZE >= 3)
dat_use1 <- subset(Homeowner_data,pick_use1)
model_logit1 <- glm(CUTENURE ~ AGE_REF + I(AGE_REF^2) + EDUC_REF + REF_RACE + I(EDUC_REFB*SEX_REF) + SEX_REF, family = binomial, data = dat_use1)
## Warning in EDUC_REFB * SEX_REF: longer object length is not a multiple of
## shorter object length
summary(model_logit1)
##
## Call:
## glm(formula = CUTENURE ~ AGE_REF + I(AGE_REF^2) + EDUC_REF +
## REF_RACE + I(EDUC_REFB * SEX_REF) + SEX_REF, family = binomial,
## data = dat_use1)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -2.3277 -0.9437 -0.7322 1.1178 1.8034
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 4.5757383 0.7190797 6.363 1.97e-10 ***
## AGE_REF -0.1757775 0.0183767 -9.565 < 2e-16 ***
## I(AGE_REF^2) 0.0018033 0.0001931 9.341 < 2e-16 ***
## EDUC_REFGrade 8 0.3206109 0.6284311 0.510 0.609928
## EDUC_REFHigh School No Degree -0.2705750 0.6041829 -0.448 0.654271
## EDUC_REFHigh School -0.8313496 0.5919403 -1.404 0.160185
## EDUC_REFSome College -1.5195697 0.5921347 -2.566 0.010280 *
## EDUC_REFAssociates -1.7119625 0.5968185 -2.868 0.004124 **
## EDUC_REFBachelors -1.8968497 0.5918159 -3.205 0.001350 **
## EDUC_REFMasters or PHD -1.9634921 0.5953018 -3.298 0.000973 ***
## REF_RACEBlack 0.5843422 0.1278914 4.569 4.90e-06 ***
## REF_RACENative American 0.8141263 0.3691101 2.206 0.027409 *
## REF_RACEAsian 0.6358050 0.1301226 4.886 1.03e-06 ***
## REF_RACEPacific Islander 2.0199211 0.7579092 2.665 0.007696 **
## REF_RACEMulti Race 0.5241916 0.2461737 2.129 0.033225 *
## I(EDUC_REFB * SEX_REF) 0.0756228 0.0644693 1.173 0.240794
## SEX_REF 0.2630499 0.0727039 3.618 0.000297 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 5119.6 on 3733 degrees of freedom
## Residual deviance: 4620.6 on 3717 degrees of freedom
## AIC: 4654.6
##
## Number of Fisher Scoring iterations: 4
nw_data2<- data.frame(AGE_REF=20:87, REF_RACE = "Black", EDUC_REF = "Bachelors", CUTENURE = "Mortgage", CUTENURE = "No Mortgage", CUTENURE = "Mortgage Status Not Reported", SEX_REF = 2)
nw_data2$yhat<-predict(model_logit1, nw_data2, type="response")
## Warning in EDUC_REFB * SEX_REF: longer object length is not a multiple of
## shorter object length
plot(nw_data2$yhat ~nw_data2$AGE , pch = 16, ylim = c(0,1.5), main = "Homeownership rate", xlab = "Age", ylab = "Percentage increase for meeting all the required variables", col = "green")
coefplot(model_logit1, innerCI=2, outerCI=0, intercept = FALSE, title = "Logit Model", color = "blue", lab = "Explantory Variables")
